Solve business challenges by using unstructured text data through Machine learning and Natural language processing.

  • Remove Duplicate Customer Records

    When stored across different data sources, customer information such as names and addresses, can be written differently, so it’s not feasible to directly match sources. Deduplication solves this challenge.

    • How can I resolve cases where the same customer is present in different data sources? See more

      Fuzzy Matching techniques match data from different sources. Additional techniques are used to perform this matching in a time-efficient manner.

  • Create Standard Product Catalogs for MDM

    Use text descriptions to map your products into a standard catalog & hierarchy.

    • How can product descriptions be used to predict a category, subcategory, brand, and sub-brand? See more

      Product Mapping and Classification uses machine-learning algorithms to classify data, such as product information, into a standard hierarchy.

  • Drive Insights from Voice of Customer Data

    Extract sentiment and key topics from product reviews, sales performance reviews, tweets, and surveys to understand what customers say and feel about your products and services.

    • How can you extract customer opinions from text? See more

      Sentiment Analysis identifies ‘sentiment carrying’ words, features, and aspects about products and brands.

    • How can you find out the most prominent topics being talked about by customers? See more

      Topic Modeling discovers topics in a collection of text documents based on their similarity.

    • What prominent aspects of your brand, product, or service are customers discussing? See more

      Context Extraction uses unsupervised text mining to identify and extract specific aspects from text data.

Case Studies

  • 2 min. read

    Leverage unstructured data to improve preventive care

    A health insurer uses unstructured data to better predict customer claims and enable focused care.

    The Big Picture A major US health insurance firm wanted to assess the riskiness of its customers. Traditionally, the company used structured data sources, such as customer demographics, past claims data, and past health...

  • 2 min. read

    Enable guided selling using natural language search queries

    A leading CPG builds a cognitive search engine to improve product recommendations.

    The Big Picture A large health and wellness products company was looking to build a cognitive learning platform that would enable guided selling based on specific consumer needs. It was also looking to create a...

  • 2 min. read

    Match customer information across multiple data sources

    A credit information services provider uses Fuzzy Matching to identify 80M more candidates to serve.

    The Big Picture One of the largest providers of credit information and information management services in the world wanted to develop an algorithm to appropriately link names and addresses across multiple datasets. The...

  • 2 min. read

    Identify social media followers’ interests to deliver targeted offers

    An analytics company uses personality identifiers to uncover targetable traits of Twitter followers.

    The Big Picture A leading loyalty analytics company wanted to identify the personality traits of ~64K of its loyalty shoppers based on their Twitter feeds. This uncovered a few challenges that needed to be examined...

  • 1 min. read

    Identify relevant report information from unstructured data

    A bank uses unstructured text to automate and expedite the cataloguing and organization of reports.

    The Big Picture A leading bank was conducting projects on socioeconomic issues, which resulted in the production of unstructured data in the form of a plethora of files. Given a database of documents and concepts as use...

Our People

  • Prashant Joshi

    Fellow

    • Leads AI & ML practice
    • Leads AI Innovation Centers for Forecasting, Text Summarization & Predictive Maintenance
    • Research and publications in Deep learning & Reservoir Computing
    • Leads AI & ML practice
    • Leads AI Innovation Centers for Forecasting, Text Summarization & Predictive Maintenance
    • Research and publications in Deep learning & Reservoir Computing
  • Amanda Dsouza

    Lead Data Scientist

    • Leads Fractal’s Text Analytics IP dCrypt and Text COE
    • Expertise in Natural Language Processing, Information Retrieval and Machine Learning
    • Leads Fractal’s Text Analytics IP dCrypt and Text COE
    • Expertise in Natural Language Processing, Information Retrieval and Machine Learning